International Journal of Engineering and Advanced Technology (IJEAT)
ISSN: 2249-8958, Volume-8, Issue- 4C, December 2018
Evaluation of Groundwater Quality at Eloor, Ernakulam
District, Kerala Using GIS
Fehmida Fatima S, Bindu A G
 cancer, poisoning by toxins etc are a few health effects of
Abstract— Pollution of groundwater resources is one of the
major challenges faced by rapidly developing countries like India.
Population explosion, large scale industrialization and
urbanization have led to rapid decline in groundwater quality.
Traditional methods of assessing groundwater quality like random
sampling and analysis has become ineffective and obsolete in
providing a comprehensive, complete and accurate picture of the
pollution levels in large areas. In this paper, groundwater quality
in the industrial belt of Eloor in Ernakulum district of Kerala is
evaluated using Geographical Information System based
geostatistical methods. Ordinary Kriging method based on spatial
autocorrelation is used for predicting the groundwater
physico-chemical parameters at unmeasured locations.
Index Terms— Geographical information system,
Geostatistical methods, Groundwater pollution, Ordinary kriging.
I. INTRODUCTION
Water is one of the most important natural resources required
for sustaining life on Earth. Water in its many forms is
required for human health, social and economic well being
and development of a nation. It is a vital resource that is
becoming increasingly scarce due to rapid increase in
population, industrialization, urbanization, climate change
and various other anthropogenic activities [9]. Water is
present on Earth in many forms; as surface water in rivers,
lakes and oceans; as snow and glacial water; as groundwater
stored in aquifers and as water vapor in atmosphere. Out of
these many sources of water, groundwater is considered as
the most dependable source because it is widely available and
reliable especially in arid climates and in times of drought
and scarcity [6]. About a third of the world’s population
depends on groundwater for meeting their basic necessities.
However, unchecked exposure to large scale pollution and
over use has lead to depletion in quality of groundwater
resources. This is encountered especially in regions of high
population density and industrialization [8].
There are many sources from which groundwater can get
polluted. These include natural sources like dissolution of
minerals found in soil and rocks and manmade sources like
leaching from landfills, application of chemical insecticides
and pesticides, industrial discharges and spills etc. The
contamination of groundwater leads to ecological
degradation, loss of vegetation, pollution of surface water
sources and causes adverse effects on human and animal
health. Once an aquifer gets contaminated, it is arduous and
unfeasible to reverse the effect. Polluted groundwater can
cause negative health impacts in humans. Kidney, liver,
nerve damages, miscarriages, developmental abnormalities,
Revised Manuscript Received on June 23, 2018.
Fehmida Fatima S, M.Tech, Environmental Engineering, Government
Engineering College Trichur, Thrissur, India, 9539026626,
Bindu A G, Civil Engineering, Government Engineering College
Trichur, Thrissur, India, 9809048663,
Retrieval Number: ICID-2018_EE_307/19©BEIESP
polluted water. Thus, better management, quality control and
monitoring is required to preserve groundwater resources [8].
Conventional methods of monitoring groundwater quality
involve construction of monitoring wells and random
sampling from bore wells or open wells. These methods are
time consuming, expensive and tedious. Alternative methods
like spatial mapping of groundwater quality using
Geographical Information System has gained momentum in
recent years. Since groundwater quality is a spatially varying
parameter, mathematical models and spatial analysis can be
used to manage groundwater problems. With limited
observations or data, manpower, expense and time, GIS
based geostatistical methods can be used as an effective tool
for obtaining accurate pollution profiles [3]-[9].
Eloor is a river island formed by the tributaries of Periyar
river in Ernakulam District of Kerala in India. It is home to
the largest industrial belt in Kerala with over 247 small,
medium and large scale industries involved in the production
of chemicals, fertilizers, insecticides, leather, rubber, metal
plating etc. Most of these industries are several decades old
and employ the most polluting manufacturing methods. This
has lead to widespread pollution of land, water and air in the
area. Eloor is considered as one of the most polluted toxic
hotspots in India by Greenpeace. A study conducted by
Greenpeace in 2003 found that pollution in Eloor has caused
harmful effects in health of local population. Eloor ranks 24 th
in the list of Critically Polluted Areas (CPAs) in the country.
There has been a significant increase in respiratory diseases,
birth defects, cancer and many other illnesses in the Eloor
community. A study conducted in 2007 on the quality of
groundwater in Eloor, found that the groundwater quality was
poor [2]. Since the detection of alarming levels of pollutants
in the wells of Eloor, a large number of people have
abandoned their wells or use it for purposes other than
drinking, cooking and bathing.
This study aims to study the status of groundwater pollution
in Eloor using GIS based geostatistical methods.
II. METHODS AND DATA
A. Study Area
Eloor is a river island occupying an area of 14.21 square
kilometers in Ernakulam District of Kerala in India between
north latitudes 9º 3´ and 10º 6´ and east longitudes 76º 20´
and 76º 28´. It is formed between two distributaries of the
Periyar river [2].
Published By:
Blue Eyes Intelligence Engineering
20 & Sciences Publication
2018 International Conference on Infrastructure Development: Issues, Innovations and the Way Forward (ICID-2018)
June 21-23, 2018 | Thiruvananthapuram, India
B. Sample Collection and Physico - Chemical Analysis 4 6.1 211 27.7 205.1
Thirty groundwater samples were collected from open 5 6.2 198 27.8 209
wells in Eloor from January 2018 to February 2018. Physico 6 5.9 185 28.2 214.2
– chemical parameters of the collected samples such as pH, 7 6.0 170 28.4 210.8
temperature, TDS, Chloride, sulphate, iron,nitrate, hardness, 8 6.9 173 28.7 223.6
COD and lead were analyzed as per APHA methods. Portable 9 6.2 198 29 246.6
pH meter and TDS meter was used to determine pH and TDS 10 6.3 230 30 243
respectively. The coordinates of sample points were obtained 11 6.2 229 28.1 245.1
using handheld GPS. 12 6.3 225 27.6 268.6
13 6.5 227 29.6 276.5
C. Geostatistical Analysis
14 6.9 216 28 267.7
Geostatistical methods such as kriging are widely used in 15 7.2 213 26 263.1
fields such as geology, mining, meteorology and earth 16 6.5 205 26.3 269.8
sciences for mapping spatial variation of parameters based on 17 6.3 219 26.6 271
spatial autocorrelation. Geostatistics uses both mathematical 18 6.1 223 28 246.4
and geostatistical methods to generate continuous surfaces. It 19 5.7 220 29.2 242.7
can be used not only to predict values at unmeasured 20 5.6 236 29.4 255.6
locations, but also to determine the uncertainty in the 21 5.2 242 28.3 304.2
prediction [2]. 22 5.6 243 28 311.6
A GIS database was created to store the analyzed water 23 5.5 238 28.2 309.9
quality parameters. Geostatistical wizard tool in GIS was 24 5.4 246 28.2 301.8
used for interpolation of parameters. Before using kriging for 25 5.9 242 27.8 283.4
interpolation, exploratory data analysis was done to 26 5.8 236 28 278.9
determine the suitability of data for prediction [1]. Kriging 27 6.4 227 27.3 248
method works best if the data is normally distributed.
28 6.1 231 29 251.5
Histograms and Q-Q plots were used to assess the normality
29 6.3 186 28 225
of the data. Parameters which did not show a normal
30 6.2 226 29 234.5
distribution were transformed using transformations such as
logarithmic and arcsine transformations to make them
conform more closely to a normal distribution. Analysis of TABLE II: ANALYSIS OF CHLORIDE, IRON , NITRATE AND
normal Q-Q plots and histograms showed that all the SULPHATE
parameters exhibited normal distribution. However,
logarithmic transformation was applied to parameters that SO42-
Sample Cl- Fe NO3-
showed a decrease in skewness after transformation. Thus, (mg/L) (mg/L) (mg/L) (mg/L)
No
pH, chloride, nitrate and sulfate data were transformed before
1 198.6 0.23 16.5 134.2
predictions.
2 202.3 0.15 19.4 136.4
Ordinary Kriging was used for prediction since it was the
3 200.1 0.32 16.8 139
easiest and most accurate of all the interpolation methods.
Semivariograms were used to study the spatial correlation in 4 210.4 0.29 15.5 145.6
data and to fit mathematical models for prediction [7]. The 5 211.9 0.27 19.3 143.2
semivariogram parameters such as nugget, sill and range 6 206.6 0.33 20.1 149.6
were studied to determine the degree of spatial correlation 7 220.8 0.18 24.3 150.3
[6]. Three types of models; spherical, exponential and 8 232.5 0.22 23.8 153.6
Gaussian were used to for prediction of parameters at 9 228.6 0.26 25 149.2
unmeasured locations. The best fitting model was selected 10 240.8 0.31 33.3 156.1
after cross-validation based on the lowest root mean square 11 252.3 0.36 38.6 163.3
error (RMSE). The smallest RMSE value indicates the most 12 228 0.33 37.3 166.1
accurate prediction [5].This model was then used to generate 13 226.1 0.3 36.5 168.8
the output surface or thematic map. 14 228.3 0.28 44.3 177.6
15 225.1 0.23 50.3 173.8
III. RESULTS AND DISCUSSIONS 16 211.5 0.25 48.9 174.2
The results of physico-chemical analysis conducted on thirty 17 213.8 0.32 52.5 176.5
groundwater samples are given in Tables I and II. The 18 224.8 0.36 48.7 173.6
parameters analyzed include pH, temperature, TDS, chloride, 19 230.2 0.35 46.2 186
nitrate, hardness, iron, and sulfate. 20 248.4 0.33 39.6 183.8
21 259.8 0.37 36.7 195.6
TABLE I: ANALYSIS OF PH, TDS, TEMPERATURE AND HARDNESS 22 260.1 0.41 29.5 204.2
Sample No pH TDS Temperatur Hardness 23 266.5 0.47 26.6 207.8
(ppm) e (mg/L) 24 271.8 0.49 24.7 211.9
˚C
1 6.7 200 29.2 186.2
2 6.2 214 28.3 194.3
3 6.0 206 28.8 211.8

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: ICID-2018_EE_307/19©BEIESP 21 & Sciences Publication
 International Journal of Engineering and Advanced Technology (IJEAT)
ISSN: 2249-8958, Volume-8, Issue- 4C, December 2018

25 261.2 0.33 27.3 216.6

26 264.8 0.38 28.3 225
27 258.6 0.3 26.6 203.6
28 238.6 0.32 24.2 223.8
29 216.2 0.28 25.6 164.6
30 235.7 0.33 39.9 145.8

Fig. 6. Best fit semivariogram model for nitrate

Semivariograms were plotted and best fitting model was
selected based on lowest RMSE value. Spherical model was
found to be best fitting model for TDS, nitrate, hardness and
iron. Gaussian model was found to be best fitting model for
temperature, chloride and sulphate. Exponential model was
found to be best fitting model for pH. The best fitting
semivariogram models for the eight groundwater quality
parameters are shown in Fig. 1 to Fig. 8. Fig. 7. Best fit semivariogram model for sulfate

Fig. 8. Best fit semivariogram model for iron

Fig. 1. Best fit semivariogram model for pH
The predicted output surface based on the best fitting models
for each of the eight parameters is shown in Fig 9 to Fig 16.

Fig. 2. Best fit semivariogram model for temperature

Fig. 3. Best fit semivariogram model for TDS Fig. 9. Predicted map of spatial variation in pH

Fig. 4. Best fit semivariogram model for hardness

Fig. 10. Prediction map of spatial variation in temperature

Fig. 5. Best fit semivariogram model for chloride

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: ICID-2018_EE_307/19©BEIESP 22 & Sciences Publication
2018 International Conference on Infrastructure Development: Issues, Innovations and the Way Forward (ICID-2018)
June 21-23, 2018 | Thiruvananthapuram, India

Fig. 11. Prediction map of spatial variation in TDS

Fig. 14. Prediction map of variation in hardness

Fig. 12. Prediction map of spatial variation in chloride Fig. 15. Prediction map of variation in sulfate

Fig. 13. Prediction map of spatial variation in nitrate Fig. 16. Prediction map of variation in iron

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: ICID-2018_EE_307/19©BEIESP 23 & Sciences Publication
 International Journal of Engineering and Advanced Technology (IJEAT)
ISSN: 2249-8958, Volume-8, Issue- 4C, December 2018

Groundwater showed high acidity in north eastern parts of
Eloor and acidity decreases towards the western and southern
regions. Temperature shows irregular variations with higher
temperature in the central, southern and eastern parts. TDS
showed higher values in the eastern region. However, TDS
was within limits in the whole area. Chloride concentration
exceeded the desirable limit of 250 mg/L in most of the
eastern parts of Eloor. Nitrate concentration was found to be
high in the north-western regions of Eloor. The desirable
limit of 45 mg/L was exceeded in this area. Hardness
exceeded the limit of 200 mg/L in almost all the regions with
the highest concentration observed in the eastern and
northern regions. Sulfate concentration was found to be
within limits except in some parts of the eastern region. The
concentration of iron exceeded the desirable limit of 0.3
mg/L in almost all regions.
From the above observations, it was inferred that the
groundwater in northern and eastern parts of Eloor are more
polluted and unfit for human consumption than the southern
regions. This variation may be attributed to the higher density
of industries in this region. Using geostatistical methods was
found to be a suitable method for predicting concentration of
groundwater parameters at unmeasured locations.
Geosci. [Online]. 7(12). pp. 5239-5252.
Available:
https://www.infona.pl/resource/bwmeta1.element.springer-ee8efe11-b
1cd-3ed2-943a-d34b091a40e3
7. M. A. Oliver and R. Webster (1990). Kriging: a method of
interpolation for geographical information systems. International
Journal of Geographical Information System. 4(3). pp. 313-332.
8. M. J. Pawari and S. Gawande. (August 2015). Water pollution and its
consequences. International Journal of Engineering Research and
General Science. [Online]. 3(4). pp. 773-776. Available:
http://pnrsolution.org/Datacenter/Vol3/Issue4/101.pdf
9. Y. Xiao, X. Gu, S.Yin, J. Shao, Y. Cui, Q. Zhang and Y. Niu.
(December 2016). Geostatistical interpolation model selection based
on ArcGIS and spatio-temporal variability analysis of groundwater
level in Piedmont Plains, Northwest China. Springer Plus. [Online].
5.pp.425-440.Available:
https://link.springer.com/article/10.1186/s40064-016-2073-0
Fehmida Fatima S was born in Thiruvananthapuram, India in 1992. She
received her B.Tech degree in Civil Engineering from College of
Engineering Thiruvananthapuram, India in 2015 and M.Tech in
Environmental Engineering from Government Engineering College Trichur,
India in 2018.
Her main areas of research interest are environmental engineering, water
treatment and groundwater studies.

ACKNOWLEDGMENT Bindu A G is an Assistant Professor in Civil

Engineering in Government Engineering College
We are grateful to Dr. Jayanand B., Principal,Govt. Trichur, India. She received her B.Tech degree in
Engineering College, Trichur, Dr. Sajikumar N., Professor Civil Engineering from Government Engineering
College Trichur, India and M.Tech in
and Head of Department, Department of Civil Engineering, Environmental Engineering from Government
Govt. Engineering College, Trichur and Dr. Vinod P., Engineering College Trichur, India.
project coordinator and PG coordinator, Department of Civil Her main areas of research interest are
Environmental forensics, Ground water contamination and Pollutant
Engineering, Govt. Engineering College, Trichur for their transport .
help and support.

REFERENCES
1. Chowdhury. (May 2016). Assessment of spatial groundwater level
variations using geostatistics and GIS in Haringhata Block, Nadia
district, West Bengal. International Journal of Research in Engineering
and Technology. [Online]. 5(5). pp. 276-280. Available:
https://ijret.org/volumes/2016v05/i05/IJRET20160505050.pdf
2. R. T. Divya, B. Sunil and C. Latha . (November 2011). Physico-
Chemical Analysis of Well Water at Eloor Industrial Area-Seasonal
Study. Current World Environment. 6(2). pp. 259-264.
3. S. Gharbia, Gharbia, S.S, T. Abushbak, H. Wafi, A. Aish, M.
Zelenakova and F. Pilla. (February 2016). Groundwater Quality
Evaluation Using GIS Based Geostatistical Algorithms. Journal of
Geoscience and Environment Protection. [Online]. 4. pp. 89-103.
Available:
https://www.scirp.org/journal/PaperInformation.aspx?PaperID=63798
4. Using ArcGIS Geostatistical Analyst, ESRI, USA, 2003.
5. H. Khosravi, K. Karimi, N. N. Fard and T. Merbahzadeh. (Winter
2016). Investigation of spatial structure of groundwater quality using
geostatistical approach in Mehran Plain, Iran. Pollution. [Online].
2(1). pp.57-65. Available:
https://jpoll.ut.ac.ir/article_55878_1e220516024e8274444f4d7f72304
cb2.pdf
6. K. Marko, A. Amri, S. Nassir and A. M. M. Elfeki (October 2013).
Geostatistical analysis using GIS for mapping groundwater quality : A
case study in the recharge area of Wadi Usfan, Saudi Arabia. Arab J

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: ICID-2018_EE_307/19©BEIESP 24 & Sciences Publication